Hmm not sure I agree here. There really shouldn't be a problem with the Album Artist field I'm not a big expert on a lot of things here, especially any MAC specific issues, but have a bunch of experience using dbPoweramp and JRiver. I have ripped a real lot of CD’s with it. JRiver does a very admirable job of automatically remapping most of the common Vobis Comment fields. I am assuming you are ripping to FLAC (most ALAC fields remap ok – but wav, mp3 etc have their limitations)
as for the field Album Artist ...
I'd look at the tag dump of your ripped files not other taggers' windows. This can be done within JRiver directly using the new tagging window or using the default tag window by clicking on the text next to the album art. You should see 2 Album Artist fields written to the file. dbPoweramp also makes a tag dump available directly from windows properties (see screen-shots). They look slightly different. Contrary to another recent post in my opinion, the fact that both are written to the file makes no difference and represents a couple of bytes extra of meta-data.
SO, for this field there should be nothing to do provided that you have correctly configured dbPoweramp to write the tags to the file. I recommend that in their meta-data options ALL tags be written (I have three exceptions ... Rating, Pre-emphasis, HDCD as I don't want their rating and the latter 2 are for tech reasons, debatable I guess).
Not meaning to go against Brian's advice, but I really strongly disagree with this setting
0. Turn off automatic tag updates. Tools > Options > General > Importing and Tagging > Update tags when file info changes > UNCHECKED
This is IMO a very important option to check. You want to insure that all your post ripping tagging is written to the file. It will only overwrite your changes. And presumably you are changing them on purpose. I now do almost all of my re-tagging in JRiver. Sometimes there are external changes done too so you want these changes to be automatically incorporated in JRiver's library too .. remember that it is possible to "tag" in JRiver without writing to the file container's themselves. Sometimes this is desirable; mostly this is not. You should check that what you consider key fields (Album Artist is a very key field) are indeed being written to the file as well as the database. There is a little flag that is checked (most default fields are already checked) "Save in file tags when possible (Options > Library & Folders > Manage Library Fields...). But when creating custom fields that are data string you should be ticking this.
Now some fields that dbPoweramp can get do need custom fields, like the sort fields as Normd's post indicates. dbPoweramp writes "Artist Sort" so the field you have to create is exactly that [Artist Sort], not [artistsort]. Bear in mind that not all cds will have all the sort fields data filled in so your mapping might be correct, there is just no metadata to transfer over.
Exception? The only mapping problem in JRiver I have had is with the Album Artist Sort field ... this is I think because it is either not present in the cd meta-data to begin with or there is something else going awry (maybe due to the special handling by JRiver of Album Artist (Auto), not sure). Anyways as Brian pointed out it is very easy to fill in semi-automatically as most of the time the Album Artist is the same as the Artist so you can just select a block of tracks and type =[Artist Sort] in the Album Artist Sort field if you choose to use one. Also, when ripping with dbPoweramp for compilation albums, making sure that Various Artist (or multiple artist, whatever you like to use) is filled in for compilation cds can avoid some unnecessary re-tagging.
At some point you'll want to write those values back to Album Artist, but I think you have to be careful because, after you create the field in step 2, it's empty inside of MC's database. If you write back to tags at that point, I'm pretty sure it will erase that field from ALL of the files. This is dangerous because you can't get it back without a restore of all the files.
For the example of a new field mapping for already imported files, you would need to run Library Tools=>Update Library from tags. Yes the newly created field is empty without doing so, but it won't instantly nuke the corresponding fields. Nevertheless, there should always be complete backups of everything just in case. For files imported afterward, it should just import the tag info automatically.
If you have the update changes set I think it is actually beneficial, rather than dangerous.
The idea being that all meta-data (tags) both in the file container itself and in the library database file within JRiver itself are, mostly, the same. You may create custom calculated fields that will not be able to be written, or if not using FLAC, some extended fields not might be able to be written to the file etc. But generally if you have a crash and need to restore, you just use the library file, fast and easy. If the library is messed up and you have no archives of these backups, then you have all really important tags already written to the file. You have 2 choices, first being Library Tools=>Update Library from Tags (note if you did not set the option, and you have not manually written the tag changes to the file by using the option Update Tags from library you are out of luck). The last option would be to re-import … again having set this option all your re-tagging. will be recoverable. However, you will nuke some information like stats, import dates etc.
One thing I learned the hard way though .. even if you set the flag to update tag information if you create a new field and choose to write the data to the file (again, recommended IMO most of the time), it will only write newly imported tag entries. You may need to run Library Tools=>Update Tags from library in certain circumstances as mentioned above. Always make a manual library backup before doing anything massive. (and make an archive backup of the JRiver backups too!)
Long story short … if JRiver is not automatically mapping Album Artist, your config either in dbPoweramp or JRiver is wrong/incomplete. If the problem is mapping more esoteric fields (other than Album Artist Sort) make sure that your custom field is exactly like the original field (including caps and/or Spaces) and it should work. Bear in mind that some fields are already mapped (eg. Performer in Vorbis Comment is Soloist if memory serves. This is more important for Picard or external taggers, rather than dbPoweramp though. If you have a habit of using external taggers, you probably should check the Update for External Changes option configured under Import.
Now I need to learn how to write shorter posts